PRACTICE GUIDELINE AASLD Guidelines for Treatment of Chronic Hepatitis B
Norah A. Terrault,1 Natalie H. Bzowej,2 Kyong-Mi Chang,3 Jessica P. Hwang,4 Maureen M. Jonas,5 and M. Hassan Murad6
1. Should adults with immune active CHB be See Editorial on Page 31 treated with antiviral therapy to decrease liver- related complications? Objectives and Guiding Principles 2. Should adults with immune-tolerant infection be treated with antiviral therapy to decrease liver- Guiding Principles related complications? This document presents official recommendations of 3. Should antiviral therapy be discontinued in hepati- the American Association for the Study of Liver Diseases tis B e antigen (HBeAg)-positive persons who have (AASLD) on the treatment of chronic hepatitis B (CHB) developed HBeAg seroconversion on therapy? virus (HBV) infection in adults and children. Unlike pre- 4. Should antiviral therapy be discontinued in per- vious AASLD practice guidelines, this guideline was sons with HBeAg-negative infection with sus- developed in compliance with the Institute of Medicine tained HBV DNA suppression on therapy? standards for trustworthy practice guidelines and uses the 5. In HBV-monoinfected persons, does entecavir ther- Grading of Recommendation Assessment, Development apy, when compared to tenofovir therapy, have a and Evaluation (GRADE) approach.1 Multiple systematic different impact on renal and bone health? reviews of the literature were conducted to support the 6. Is there a benefit to adding a second antiviral agent recommendations in this practice guideline. An enhanced in persons with persistent low levels of viremia while understanding of this guideline will be obtained by read- being treated with either tenofovir or entecavir? ing the applicable portions of the systematic reviews. 7. Should persons with compensated cirrhosis and low This guideline focuses on using antiviral therapy in levels of viremia be treated with antiviral agents? chronic HBV infection and does not address other related 8. Should pregnant women who are hepatitis B sur- and important issues, such as screening, prevention, and face antigen (HBsAg) positive with high viral surveillance. For broader issues related to diagnosis, sur- load receive antiviral treatment in the third tri- veillance, and prevention as well as treatment in special mester to prevent perinatal transmission of HBV? populations (e.g., liver transplant recipients) that are not 9. Should children with HBeAg-positive CHB be addressed by this guideline, the previous AASLD guide- treated with antiviral therapy to decrease liver- line2 and recent World Health Organization (WHO) related complications? 3 guideline are excellent additional resources. Target Audience Objectives This guideline is intended primarily for health care Guideline developers from the AASLD formulated a professionals caring for patients with CHB. Addition- list of discrete questions that physicians are faced with ally, this guideline may assist policy makers in optimiz- in daily practice. These questions were: ing the care of individuals living with CHB.
Abbreviations: AASLD, American Association for the Study of Liver Diseases; ALT, alanine aminotransferase; anti-HBe, antibody to HBeAg; anti-HBs, antibody to HBsAg; APRI, AST-to-platelet ratio index; AST, aspartate aminotransferase; cccDNA, covalently closed circular DNA; CHB, chronic hepatitis B; CI, confidence interval; GRADE, Grading of Recommendation Assessment, Development and Evaluation; HBV, hepatitis B virus; HCV, hepatitis C virus; HCC, hepatocellular carci- noma; HIV, human immunodeficiency virus; HDV, hepatitis delta virus; HBsAg, hepatitis B surface antigen; HBeAg, hepatitis B e antigen; IFN, interferon; NA, nucle- os(t)ide analog; Peg-IFN, pegylated interferon; RCT, randomized, controlled trial; RR, relative risk; ULNs, upper limits of normal; WHO, World Health Organization. From the 1University of California San Francisco, San Francisco, CA; 2Ochsner Medical Center, New Orleans, LA; 3Corporal Michael J. Crescenz VA Medical Center & University of Pennsylvania Perelman School of Medicine, Philadelphia, PA; 4The University of Texas MD Anderson Cancer Center, Houston, TX; 5Bos- ton Children’s Hospital, Harvard Medical School, Boston, MA; 6Mayo Clinic, Rochester, MN. Received August 24, 2015; accepted August 25, 2015. Additional Supporting Information may be found at onlinelibrary.wiley.com/doi/10.1002/hep.28156/suppinfo. The funding for the development of this Practice Guideline was provided by the American Association for the Study of Liver Diseases. This Practice Guideline was approved by the AASLD on August 1, 2015. This Practice Guideline published with accompanying Reviews by Lok et al., Jonas et al., and Brown et al.
261 262 TERRAULT ET AL. HEPATOLOGY, January 2016
Table 1. Phases of CHB Infection
ALT HBV DNA HBeAg Liver Histology
Immune-tolerant phase Normal Elevated, typically Positive Minimal inflammation and fibrosis >1 million IU/mL HBeAg-positive Elevated Elevated Positive Moderate-to-severe inflammation or fibrosis immune-active phase 20,000 IU/mL Inactive CHB phase Normal Low or undetectable Negative Minimal necroinflammation but variable fibrosis <2,000 IU/mL HBeAg-negative immune Elevated Elevated Negative Moderate-to-severe inflammation or fibrosis reactivation phase 2,000 IU/mL
need for treatment. Of note, some persons will be in the Background “gray zones,” meaning that their HBV DNA and ALT Burden of Disease levels do not fall into the same phase. Longitudinal Globally, an estimated 240 million persons have follow-up of ALT and HBV DNA levels and/or assess- CHB with a varying prevalence geographically, highest ment of liver histology can serve to clarify the phase of in Africa and Asia.4 In the United States, the National infection. Health and Nutrition Examination Survey (1999 to i. Immune-tolerant phase: In this highly replicative/ 2008) identified approximately 704,000 adults with low inflammatory phase, HBV DNA levels are 5 CHB, but with adjustments for hepatitis B infection elevated, ALT levels are normal (<19 U/L for among foreign-born persons, the upper estimate of females and <30 U/L for males), and biopsies CHB in the United States may be as high as 2.2 are without signs of significant inflammation or 6 million. Globally, deaths from cirrhosis and hepatocel- fibrosis. The duration of this phase is highly vari- lular carcinoma (HCC) were estimated at 310,000 and able, but longest in those who are infected peri- 340,000 per year, respectively.7 To reduce the morbidity natally. With increasing age, there is an and mortality of CHB in the United States and world- increased likelihood of transitioning from wide, there is a need for continued efforts to identify immune-tolerant to the HBeAg-positive immune- infected individuals through targeted screening, prevent active phase. new infections through vaccination, and monitor and ii. HBeAg-positive immune-active phase: Elevated treat those at risk for complications of their CHB, ALT and HBV DNA levels in conjunction with including surveillance for HCC.8,9 liver injury characterize this phase. Median age of onset is 30 years among those infected at a young Natural History in Adults and Children age. The hallmark of transition from the HBeAg- CHB has been traditionally characterized into four positive immune-active to -inactive phases is phases (Table 1), reflecting the dynamic relationship HBeAg seroconversion. The rate of spontaneous between viral replication and evolution and the host seroconversion from HBeAg to antibody to HBeAg immune response. These phases are of variable duration (anti-HBe) is less than 2% per year in children and not every person infected with CHB will evolve younger than 3 years of age and increases during through all phases. Given the dynamic nature of CHB puberty and among adults to 8% and 12% per infection, serial monitoring of HBV DNA and alanine year, respectively. aminotransferase (ALT) levels is important to character- iii. Inactive CHB phase: In this phase, HBV DNA ize the phase of infection. A single ALT and HBV DNA levels are low or undetectable, ALT levels are nor- level are insufficient to assign phase of infection and/or mal, and anti-HBe is present. Liver histology
Address reprint requests to: Norah Terrault, M.D., M.P.H., Division of Gastroenterology, University of California San Francisco, 513 Parnassus Avenue, S357, San Francisco, CA 94143-0538. E-mail: [email protected]; fax: 11-415-502-6714. Copyright VC 2015 by the American Association for the Study of Liver Diseases. View this article online at wileyonlinelibrary.com. DOI 10.1002/hep.28156 Potential conflict of interest: Dr. Jonas consults and received grants from Gilead. She received grants from Bristol-Myers Squibb and Roche. Dr. Chang advises Genentech, Alnylam, and Arbutus. Dr. Terrault consults for Bristol-Myers Squibb and received grants from Gilead. Dr. Bzowej received grants from Gilead, Synageva, and Ocera. HEPATOLOGY, Vol. 63, No. 1, 2016 TERRAULT ET AL. 263
Table 2. Host, Viral/Disease, and Environmental Factors region, and liver histology shows necroinflamma- Associated With Cirrhosis and HCC tion and fibrosis. Persons with HBeAg-negative Cirrhosis HCC CHB tend to have lower serum HBV DNA levels Host >40 years of age >40 years of age than those with HBeAg-positive CHB and are more Male sex Male sex likely to experience a fluctuating course. Immune compromised Immune compromised Positive family history Resolved CHB infection is defined by clearance of Born in Sub-Saharan Africa HBsAg with acquisition of antibody to HBsAg. Approx- Viral/ High serum HBV DNA Presence of cirrhosis disease (>2,000 IU/mL) High serum HBV DNA imately 0.5% of persons with inactive CHB will clear Elevated ALT levels (>2,000 IU/mL) HBsAg yearly; most will develop antibody to HBsAg Prolonged time to HBeAg Elevated ALT (anti-HBs). Low levels of HBV DNA are transiently seroconversion Prolonged time to HBeAg detected in the serum in the minority of persons achiev- Development of seroconversion 10,11 HBeAg-negative CHB Development of ing seroclearance. Clearance of HBsAg, whether Genotype C HBeAg-negative CHB spontaneous or after antiviral therapy, reduces risk of Genotype C hepatic decompensation and improves survival. Environmental Concurrent viral infections Concurrent viral infections (HCV, HIV, and HDV) (HCV, HIV, and HDV) Risk of liver-related complications is variable. Among Heavy alcohol use Heavy alcohol use untreated adults with CHB, cumulative 5-year incidence Metabolic syndrome Metabolic syndrome of cirrhosis is 8%-20%, and among those with cirrhosis, (obesity, diabetes) (obesity, diabetes) Aflatoxin 5-year cumulative risk of hepatic decompensation is 20%, 12-14 Smoking and risk of HCC is 2%-5%. Viral, host, and environ- mental factors influence risks of cirrhosis and HCC13 (Table 2). HBV DNA levels, ALT levels, and HBeAg sta- tus are among the most important determinants of risk of shows minimal necroinflammation, but variable progression to cirrhosis,15,16 whereas HBV DNA levels fibrosis reflecting previous liver injury during the (>2,000 IU/mL), HBeAg status, and cirrhosis are key HBeAg-positive immune-active phase. Among per- predictors of HCC risk.15-18 A biological gradient of risk sons who undergo spontaneous HBeAg seroconver- has been shown in adults with HBV DNA levels above sion, 67%-80% will continue to remain in the 2,000 IU/mL; a higher HBV DNA level is associated inactive CHB phase. Approximately 4%-20% of with progressively higher rates of cirrhosis and HCC.15 inactive carriers have one or more reversions back to HBeAg positive. Diagnosis, Staging and Monitoring of Persons iv. HBeAg-negative immune reactivation phase: With CHB Among those who seroconvert from HBeAg to anti- The initial evaluation of persons with CHB should HBe positive, 10%-30% continue to have elevated include a thorough history and physical examination, ALT and high HBV DNA levels, and roughly 10%- with special emphasis on risk factors for coinfection, alco- 20% of inactive carriers may have reactivation of hol use, and family history of HBV infection and liver HBV replication and exacerbations of hepatitis cancer. Laboratory tests should include assessment of liver after years of quiescence. Most of these persons har- disease activity and function, markers of HBV replication, bor HBV variants in the precore or core promoter and tests for coinfection with hepatitis C virus (HCV),
Table 3. Initial Evaluation of HBsAg-Positive Patient
History/Physical Examination Routine Laboratory Tests Serology/Virology Imaging/Staging Studies
All patients Symptoms/signs of cirrhosis CBC including platelet count, HBeAg/anti-HBe Abdominal ultrasound Alcohol and metabolic risk factors AST, ALT, total bilirubin, HBV DNA quantitation Vibration-controlled transient Family history of HCC alkaline phosphatase, Anti-HAV to determine need elastography or serum fibrosis Vaccination status albumin, INR for vaccination panel (APRI, FIB-4, or FIbroTest) Select patients Tests to rule out other causes of HBV genotype Liver biopsy chronic liver diseases if Anti-HDV elevated liver test(s) Anti-HCV AFP, GGT Anti-HIV in those who have not undergone one-time screening (ages 13-64)
Abbreviation:s INR, international normalized ratio; GGT, gamma-glutamyl transpeptidase. 264 TERRAULT ET AL. HEPATOLOGY, January 2016 hepatitis delta virus (HDV), or human immunodeficiency surveillance. Liver biopsy provides an assessment of the virus (HIV) in those at risk (Table 3). Owing to the fluctu- severity of necroinflammation and fibrosis, rules out ating nature of CHB, the accuracy of one high HBV other causes of liver disease, and may be especially useful DNA level at a single time point in predicting prognosis is for persons who lack clear-cut indications for treatment. poor and regular monitoring of disease status is imperative Whereas liver biopsy is regarded as the best method to to determine need for antiviral therapy. The upper limits assess the severity of inflammatory activity and fibrosis, of normal (ULNs) for ALT values based on healthy sub- noninvasive methods to assess fibrosis severity are also jects are lower than laboratory values derived from all pop- useful. Acute-on-chronic exacerbations of hepatitis B ulations, including those with subclinical liver disease.19 may lead to overestimation of fibrosis stage by noninva- Determination of the stage of liver disease is impor- sive tests, and different cutoffs for significant and tant in guiding antiviral therapy decisions and need for advanced fibrosis depending on ALT levels have been
Table 4. Approved Antiviral Therapies in Adults and Children
Pregnancy Drug Dose in Adults* Use in Children* Category Potential Side Effects† Monitoring on Treatment†
Peg-IFN-2a(adult) 180 lg weekly 1 year C Flu-like symptoms, fatigue, CBC (monthly to every 3 months) IFN-a-2b Dose: 6 million IU/m2 mood disturbances, cytope- TSH (every 3 months) (children) TIW‡ nias, autoimmune disorders Clinical monitoring for autoimmune, in adults ischemic, neuropsychiatric, and Anorexia and weight loss in infectious complications children Lamivudine 100 mg daily 2 years C Pancreatitis Amylase if symptoms Dose: 3 mg/kg daily to Lactic acidosis Lactic acid levels if clinical concern max 100 mg Telbivudine 600 mg daily — B Creatine kinase elevations and Creatine kinase if symptoms myopathy Cinical evaluation if symptoms Peripheral neuropathy Lactic acid levels if clinical concern Lactic acidosis Entecavir 0.5 or 1.0 mg daily§ 2 years C Lactic acidosis Lactic acid levels if clinical concern Dose: weight-based to 10- 30 kg; above 30 kg 0.5 mg dailyk Adefovir 10 mg daily 12 years 10 mg daily C Acute renal failure Creatinine clearance at baseline Fanconi syndrome If at risk for renal impairment, creati- Nephrogenic diabetes insipidus nine clearance, serum phosphate, Lactic acidosis urine glucose, and protein at least annually Consider bone density study at base- line and during treatment in per- sons with history of fracture or risks for osteopenia Lactic acid levels if clinical concern Tenofovir 300 mg daily 12 years 300 mg daily B Nephropathy, Fanconi syndrome Creatinine clearance at baseline Osteomalacia If at risk for renal impairment, creati- Lactic acidosis nine clearance, serum phosphate, urine glucose, and protein at least annually Consider bone density study at base- line and during treatment in per- sons with history of fracture or risks for osteopenia Lactic acid levels if clinical concern
*Doses need to be adjusted in persons with renal dysfunction. †Per package insert. ‡Peg-IFN-a-2a is not approved for children with CHB, but is approved for treatment of chronic hepatitis C. Providers may consider using this drug for children with chronic HBV. The duration of treatment indicated in adults is 48 weeks. §Entecavir dose in adults is 1 mg daily if lamivudine or telbivudine experienced or decompensated cirrhosis. kEntecavir doses in treatment-na€ıve children older than 2 and at least 10 kg are: 0.15 mg (10-11 kg), 0.2 mg (>11-14 kg), 0.25 mg (>14-17 kg), 0.3 mg (>17-20 kg), 0.35 mg (>20-23 kg), 0.4 mg (>23-26 kg), 0.45 mg (>26-30 kg), and 0.5 mg (>30 kg). For treatment-experienced children older than 2 and at least 10 kg, the entecavir doses are: 0.30 mg (10-11 kg), 0.4 mg (>11-14 kg), 0.5 mg (>14-17 kg), 0.6 mg (>17-20 kg), 0.7 mg (>20-23 kg), 0.8 mg (>23-26 kg), 0.9 mg (>26-30 kg), and 1.0 mg (>30 kg). Abbreviations: CBC, complete blood counts; TSH, thyroid-stimulating hormone. HEPATOLOGY, Vol. 63, No. 1, 2016 TERRAULT ET AL. 265 proposed.20 Serum markers of fibrosis, such as aspartate frequent with interferon (IFN) therapy than with aminotransferase (AST)-to-platelet ratio index (APRI), nucleos(t)ide analogs (NAs) therapy. Overall, all NAs have FIB-4, FibroTest, and vibration-controlled transient an excellent safety profile across a wide spectrum of per- elastography, have only moderate accuracy in identifying sons with CHB, including those with decompensated cir- persons with significant fibrosis (fibrosis stage 2 or rhosis and transplant recipients.25 The side effects listed in greater on the Metavir scale), but good diagnostic accu- Table4forNAsareinfrequent.ForpersonswithHDV racy in excluding advanced fibrosis21,22 and may be use- coinfection, the only effective treatment is pegylated inter- ful aids in decision making. feron (Peg-IFN). For persons with HIV coinfection, treat- ment of HBV needs to be coordinated with HIV therapy Antiviral Therapy given that several HBV drugs have anti-HIV activity 26 The goals of antiviral treatment are to decrease the (tenofovir, entecavir, lamivudine, and telbivudine). morbidity and mortality related to CHB. The achieve- Biochemical, serological, virological, and histological ment of a sustained suppression of HBV replication has endpoints are used to assess the success of therapy (Table 5). been associated with normalization of serum ALT, loss Assessments are performed on continuous therapy 27-30 of HBeAg with or without detection of (anti-HBe), and (NAs) and after therapy discontinuation (Peg- 2,31,32 improvement in liver histology. Historically, the term IFN). The best predictor of sustained remission “cure” was avoided in treatment of CHB, given that per- off-treatment is HBsAg loss, but this is infrequently sistence of covalently closed circular DNA (cccDNA), achieved with current therapies. the transcriptional template of HBV,23,24 in the nucleus of hepatocytes, even in persons with serological markers Methods of Guideline Development of resolved infection, poses a lifelong risk for reactiva- The specific questions specified a priori for evaluation tion of infection. However, an immunological cure may by the guidelines committee are shown in Table 6. be defined by HBsAg loss and sustained HBV DNA A methodologist moderated and facilitated the pro- suppression and a virological cure defined by eradication cess of question development. A separate group of of virus, including the cccDNA form. The latter is not AASLD content experts collaborated with an independ- currently an attainable goal. ent research group with expertise in conducting system- There are six therapeutic agents approved for the atic reviews to synthesize the available evidence treatment of adults with CHB in the United States and informing these key questions. By multiple face-to-face five therapeutic agents approved for the treatment of meetings, phone conferences, and electronic communi- children with CHB (Table 4). Side effects are more cation, the systematic review group finalized evidence
Table 5. Efficacy of Approved Preferred Antiviral Therapies in Adults With Treatment-Na€ıve CHB and Immune Active Disease (Not Head-to-Head Comparisons)
Peg-IFN* (%) Entecavir† (%) Tenofovir† (%)
HBeAg-Positive HBV DNA suppression‡ 30-42 (<2,000-40,000 IU/mL) 61 (<50-60 IU/mL) 76 (<60 IU/mL) 8-14 (<80 IU/mL) HBeAg loss 32-36 22-25 — HBeAg seroconversion 29-36 21-22 21 Normalization ALTk 34-52 68-81 68 HBsAg loss 2-7 (6 mos post-treatment) 2-3 (1 yr) 3 (1 yr) 11 (at 3 yrs post-treatment) 4-5 (2 yrs) 8 (3 yrs) (References) 31,33-35 36-38 30,39 HBeAg-Negative HBV DNA suppression§ 43 (<4,000 IU/mL) 90-91 93 19 (<80 IU/mL) Normalization ALTk 59 78-88 76 HBsAg loss (%) 4 (6 mos post-treatment) 0-1 (1 yr) 0 (1 yr) 6 (at 3 yrs post-treatment) (References) 40,41 42 39
*Assessed 6 months after completion of 12 months of therapy. †Assessed after 2-3 years of continuous therapy. ‡HBV DNA <2,000-40,000 IU/mL for Peg-IFN; <60 IU/mL for entecavir and tenofovir. §HBV DNA <20,000 IU/mL for Peg-IFN; <60 IU/mL for entecavir and tenofovir. kALT normalization defined by laboratory normal. 266 TERRAULT ET AL. HEPATOLOGY, January 2016
Table 6. Clinical Questions Evaluated
Question Population Intervention Comparison Outcome(s)
1 Immune-active CHB Antiviral therapy No treatment Cirrhosis, decompensation, HCC, death, loss of HBsAg 2 Immune-tolerant CHB, adults Antiviral therapy No treatment Cirrhosis, decompensation, HCC, death, loss of HBsAg 3 HBeAg-positive immune-active chronic Continued antiviral Stopping antiviral Cirrhosis, HCC, reactivation, seroreversion, hepatitis, with HBeAg seroconversion therapy therapy decompensation, loss of HBsAg on therapy 4 HBeAg-negative immune-active chronic Continued antiviral Stopping antiviral Reactivation, decompensation, loss of hepatitis, with viral suppression on therapy therapy HBsAg antiviral therapy 5 CHB on treatment with oral therapy Tenofovir Entecavir Renal function, hypophosphatemia, bone health 6 CHB on treatment with oral therapy Continue therapy Change or switch HBV resistance, clinical flare, decompen- with persistent viremia therapy sation, loss of HBeAg 7 CHB with cirrhosis, with HBV DNA Antiviral therapy No treatment Decompensation, HCC, death <2,000 IU/mL 8 Pregnant women with CHB Antiviral therapy in No treatment CHB in the infant, maternal safety, fetal/ third trimester infant safety 9 HBeAg-positive CHB, children/ Antiviral therapy No treatment Cirrhosis, decompensation, HCC, death, adolescents HBeAg seroconversion, loss of HBsAg summaries following the GRADE approach (Table 7).1 of the key questions are presented as an appendix to this In this approach, the quality of evidence (i.e., certainty article. For the remaining questions with sparse and in evidence) is rated as high, moderate, low, or very low indirect evidence, relevant studies are summarized after based on the domains of precision, directness, consis- each recommendation. tency, and risk of bias and publication bias. The guideline-writing group based its recommendations on the quality of evidence, balance of benefits and harms, Treatment of Persons With Immune-Active patients’ values and preferences, and clinical context. CHB Recommendations are graded as strong (apply to most Recommendations patients with minimal variation) or conditional (apply to the majority of patients whose values and preferences 1A. The AASLD recommends antiviral therapy for are consistent with the course of action). Technical adults with immune-active CHB (HBeAg negative or remarks are added to recommendations to facilitate HBeAg positive) to decrease the risk of liver-related implementation. Evidence profiles corresponding to five complications.
Table 7. The GRADE Approach 1. Rating the Quality of Evidence Study design Initial rating of quality of evidence Rate down when Rate up when RCT High Risk of bias Large effect (e.g., RR: 0.5) Moderate Inconsistency Very large effect (e.g., RR: 0.2) Imprecision Dose response gradient Observational Low Indirectness All plausible confounding would increase Very low Publication bias the association 2. Determinants of the Strength of a Recommendation Quality of evidence Balance of benefits and harms Patient values and preferences Resources and costs 3. Implications of the Strength of Recommendation Strong Population: Most people in this situation would want the recommended course of action and only a small proportion would not. Health care workers: Most people should receive the recommended course of action. Policy makers: The recommendation can be adapted as a policy in most situations. Conditional Population: The majority of people in this situation would want the recommended course of action, but many would not. Health care workers: Be prepared to help patients make a decision that is consistent with their values using decision aids and shared decision making. Policy makers: There is a need for substantial debate and involvement of stakeholders. HEPATOLOGY, Vol. 63, No. 1, 2016 TERRAULT ET AL. 267
Quality/Certainty of Evidence: Moderate ance with long-term use. Patient-specific factors Strength of Recommendation: Strong that need to be considered in choosing between Peg- IFN, entecavir, and tenofovir include: 1B. The AASLD recommends Peg-IFN, entecavir, or tenofovir as preferred initial therapy for adults Desire for finite therapy (see below) with immune-active CHB. Anticipated tolerability of treatment side effects (Table 4). Quality/Certainly of Evidence: Low Comorbidities: Peg-IFN is contraindicated in Strength of Recommendation: Strong persons with autoimmune disease, uncon- Technical Remarks trolled psychiatric disease, cytopenias, severe cardiac disease, uncontrolled seizures, and 1. Immune-active CHB is defined by an elevation decompensated cirrhosis. of ALT >2 ULN or evidence of significant histo- Previous history of lamivudine resistance logical disease plus elevated HBV DNA above (entecavir is not preferred in this setting). 2,000 IU/mL (HBeAg negative) or above Family planning: A finite therapy with Peg- 20,000 IU/mL (HBeAg positive). IFN pre-pregnancy or use of oral antiviral 2. The ULN for ALT in healthy adults is 30 U/L that is safe in pregnancy is best (Table 4). for males and 19 U/L for females. HBV genotype: A and B genotypes are more 3. There is insufficient evidence for or against use of likely to achieve HBeAg and HBsAg loss with ALT criterion other than ALT 2ULN.Thedeci- Peg-IFN than non-A/B genotypes. sion to treat persons with ALT above the ULNs, Medication costs. but <2 ULN, requires consideration of severity of 7. Peg-IFN is preferred over nonpegylated forms liver disease (defined by biopsy or noninvasive for simplicity. testing). Therapy is recommended for persons 8. For persons treated with Peg-IFN, 48 weeks dura- with immune-active CHB and cirrhosis if HBV tion is used in most studies and is preferred. This DNA >2,000 IU/mL, regardless of ALT level. treatment duration yields HBeAg seroconversion 4. Additional factors included in the decision to rates of 20%-31%31 and sustained off-treatment treat persons with immune-active CHB but ALT HBV DNA suppression <2,000 IU/mL in 65% <2 ULN and HBV DNA below thresholds are: of persons who achieve HBeAg to anti-HBe sero- Age: Older age (>40 years) is associated with conversion.32 The combination of Peg-IFN and higher likelihood of significant histological NAs has not yielded higher rates of off-treatment disease. serological or virological responses and is not Family history of HCC recommended.43 Previous treatment history: 9. Duration of therapy for NA-based therapy is - Serological benefits of Peg-IFN (HBeAg and variable and influenced by HBeAg status, dura- HBsAg loss) may occur months to years after tion of HBV DNA suppression, and presence of treatment discontinuation (delayed). cirrhosis/decompensation. All NAs require dose - Previous NA exposure is a risk for drug resistance adjustment in persons with creatinine clearance Presence of extrahepatic manifestations: Indi- <50 mL/min. cation for treatment independent of liver dis- 10. Evaluation for stage of disease using noninva- ease severity sive methods or liver biopsy is useful in guiding treatment decisions including duration of 5. Level of HBV DNA should be compatible with therapy. immune-active disease and the cutoffs recom- 11. Treatment with antivirals does not eliminate mended should be viewed as a sufficient, but the risk of HCC, and surveillance for HCC not absolute, requirement for treatment. should continue in persons who are at risk. 6. Head-to-head comparisons of antiviral therapies fail to show superiority of one therapy over another Background in achieving risk reduction in liver-related compli- CHB is a dynamic disease characterized by variable peri- cations. However, in recommending Peg-IFN, teno- ods of immune activity versus quiescence that culminates fovir, and entecavir as preferred therapies, the most in the development of cirrhosis, liver cancer, and liver- important factor considered was the lack of resist- related death in a proportion of persons. Elevated serum 268 TERRAULT ET AL. HEPATOLOGY, January 2016
ALT and HBV DNA levels are strongly predictive of type of therapy, IFN and NAs achieved long-term bene- risk of liver complications.15,16 Other factors include fits of preventing cirrhosis and HCC, but only NAs older age, male sex, a family history of HCC, alcohol were associated with reduced rates of decompensation use, HIV infection, diabetes, HBV genotype C, and and death. HBV precore and core promoter variants. The goal of The primary indication for treatment initiation in a HBV therapy is to prevent liver-related morbidity and person with immune-active disease is the presence of sig- mortality. Persons in the immune-active phases of infec- nificant liver injury or fibrosis, as reflected by elevated tion (HBeAg positive and negative) display elevated ALT levels or moderate-to-severe necroinflammatory ALT, histological evidence of liver injury (significant activity on histology and/or fibrosis plus active HBV inflammation and/or fibrosis), and elevated HBV DNA viremia. Clinical trials of treatment in adults used labo- levels with a greater risk of progressive liver disease and ratory ULNs for ALT to define elevated ALT and typi- its associated complications. cally required ALT elevation 1.3-2.0 times ULNs for inclusion. It is recognized that the normal ALT levels of Evidence and Rationale healthy adults are 30 U/L for males and 19 U/L for The evidence profile is summarized in Supporting females.19 Thus, using these ALT cutoffs for normal, the Table 1.44 A total of 42 studies were included comparing recommendation to consider treatment of adults with antivirals to no treatment and reporting outcomes of cir- ALT values of 2 times the ULN (>60 U/L for males rhosis, HCC, decompensation, or death. Seven studies and >38 U/L for females) is more inclusive than the were randomized, controlled trials (RCTs) and 35 studies ALT criteria used in the clinical trials. The HBV DNA were observational; a total of 13 studies provided out- levels used to define immune-active disease are based on comes in persons with cirrhosis, and two studies provided historical cutoffs of clinical trials, with supportive evi- outcomes in persons with decompensated cirrhosis. dence from natural history studies showing that the rela- Regarding specific antiviral therapies, 16 studies com- tive risk of liver-related complications increases with pared IFN to no treatment and 27 studies compared NA HBV DNA levels above 2,000 IU/mL.15,16 In our sys- therapy to no treatment. A network meta-analysis to tematic review, three studies comparing liver-related compare antiviral therapies was not feasible owing to the outcomes in persons receiving antiviral therapy versus small number of RCTs per analysis. The quality of evi- control stratified by HBV DNA level (<2,000 vs. dence was generally higher for RCTs (range, very low to >2,000 IU/mL) and found no significant difference in high; majority, low to moderate) versus observational outcomes. studies (very low). Number of RCTs (range, 1-6 per out- Liver biopsies are not required to make treatment come) was lower than observational studies (1-23 per out- decisions. However, determination of the presence of come). For specific NAs, the number of studies was advanced fibrosis previous to treatment is important in limited and quality highly variable. The magnitude of the guiding treatment choices, duration of therapy, and treatment effect (40%-61% reduction in liver-related therapeutic endpoints. Available evidence does not complications: cirrhosis, decompensation, HCC, and define the specific ALT and HBV DNA thresholds at death) and consistency of risk reduction across studies which treatment should be initiated. A high baseline and among subgroups contributed to strength of the rec- ALT, 2-5 times ULN (based on laboratory ULN), and ommendation despite lower quality of the studies. moderate-to-high necroinflammatory activity on biopsy Antiviral therapy (compared to no treatment) was are associated with higher likelihood of achieving the associated with significant risk reductions in cirrhosis in intermediate outcomes with treatment (HBeAg serocon- observational studies (relative risk [RR] 5 0.39; 95% version and HBV DNA <2000 IU/mL post-treatment). confidence interval [CI]: 0.20-0.75) and RCTs (RR 5 Noninvasive tests, such as elastography, may be useful in 0.55; 95% CI: 0.38-0.78). Observational studies (n 5 ruling out cirrhosis (i.e., have high negative predictive 23) showed a risk reduction in HCC (RR 5 0.49; 95% value), but are less accurate in predicting presence of sig- CI: 0.35-0.70) and death (RR 5 0.6; 95% CI: 0.5-0.8) nificant fibrosis (F2 or higher). High necroinflammatory and RCTs showed a risk reduction in decompensation activity and high ALT levels are associated with (RR 5 0.44; 95% CI: 0.29-0.68). Among the subgroup increased stiffness and this needs to be taken into con- of persons with cirrhosis, antiviral therapy (vs. no treat- sideration in interpreting results.45 ment) yielded risk reductions of HCC (RR 5 0.54; 95% CI: 0.41-0.72) and decompensated liver disease Future Research (RR 5 0.45; 95% CI: 0.22-0.89), but not in mortality Future studies are needed to better define risk benefit (RR 5 0.68; 95% CI: 0.40-1.18). In assessment by for treating persons with mild ALT elevation (e.g. 1-2 3 HEPATOLOGY, Vol. 63, No. 1, 2016 TERRAULT ET AL. 269
ULN) and low-level HBV DNA (e.g., <20,000 IU/mL tolerant phase of infection would benefit from antiviral for HBeAg positive and <2,000 IU/mL for HBeAg nega- therapy. Of note, these natural history studies used tive) who are currently in the “gray zone” for ALT and ALT<45 U/L as ULNs. In cross-sectional studies using HBV DNA criteria for treatment versus observation. more-stringent ALT criteria of 30 U/L for males and Studies to define the use of noninvasive measures of dis- 19 U/L for females, significant histological disease ease severity in treatment algorithms are important. There (fibrosis 2/4 and necroinflammatory score 2/4) is is also a great need for newer treatment approaches that found in the minority ( 20%) of HBeAg-positive eliminate the HBV cccDNA to achieve virological cure. adults with high HBV DNA (>106 IU/mL).46,47 In persons who acquired their infection at birth or in early Treatment of Adults With childhood, the average age of transitioning from Immune-Tolerant CHB immune-tolerant to immune-clearance phases is 30 years.47 Age over 40 years is associated with higher like- Recommendations lihood of significant histological disease in HBeAg- 46,48 2A. The AASLD recommends against antiviral positive persons with normal ALT levels. therapy for adults with immune-tolerant CHB. Evidence and Rationale Quality/Certainly of Evidence: Moderate The evidence profile is summarized in Supporting Strength of Recommendation: Strong Table 2. Among 17 studies of interventions in immune- tolerant adults, only two examined adults with ALT less Technical Remark than ULNs, whereas most used ALT less than 2 times 1. Immune-tolerant status should be defined by ALT ULNs for inclusion. All were RCTs with treatment dura- levels utilizing 30 U/L for men and 19 U/L tion of 24-48 weeks for IFN or 48 weeks for NAs with 6- for women as ULNs rather than local laboratory 12 months of post-treatment follow-up. All studies used ULNs. HBeAg loss and seroconversion as the primary endpoint, whereas only two studies evaluated HBsAg loss. Five 2B. The AASLD suggests that ALT levels be tested studies comparing antiviral therapy to placebo/no treat- at least every 6 months for adults with immune- ment were the primary studies informing this recommen- tolerant CHB to monitor for potential transition to dation. The remaining 12 studies were head-to-head immune-active or -inactive CHB. comparisons of different antiviral therapies. Quality/Certainly of Evidence: Very low Compared to untreated/placebo controls, any antivi- Strength of Recommendation: Conditional ral therapy resulted in a significantly higher rate of HBeAg loss (RR, 2.69; 95% CI: 1.19-6.09) and sero- 2C. The AASLD suggests antiviral therapy in the conversion (RR, 2.22; 95% CI: 1.2-4.09). Stratification > select group of adults 40 years of age with normal of results by treatment type (IFN and NAs, all lamivu- ALT and elevated HBV DNA ( 1,000,000 IU/mL) dine) yielded RR that included 1 (not significantly dif- and liver biopsy showing significant necroinflamma- ferent from untreated controls). The RCT studies were tion or fibrosis. low-to-moderate quality and the RCTs limited to per- Quality/Certainly of Evidence: Very low sons with baseline ALT values less than ULNs were very Strength of Recommendation: Conditional low to low quality. There are no studies demonstrating that antiviral Technical Remark therapy is beneficial in reducing rates of HCC, cirrhosis, and liver-related death in persons with immune-tolerant 1. Moderate-to-severe necroinflammation or fibrosis CHB. Finite treatment duration for 24-48 months on liver biopsy is a reason to consider initiation yields higher rates of HBeAg seroconversion, but not of antiviral therapy, if other causes of liver dis- HBsAg seroconversion, and only among studies includ- ease are excluded. ing persons with ALT <2 ULN. The latter group likely Background included persons with HBeAg-positive immune active Natural history studies have found a strong associa- disease, a group recommended for antiviral therapy. tion between serum HBV DNA levels and the develop- Given the lack of evidence of benefit to those with ALT ment of HCC and cirrhosis, independent of serum ALT